Public professionals and policy implementation:

conceptualizing and measuring three types of role

conflicts

Lars Tummers, Brenda Vermeeren, Bram Steijn and Victor Bekkers

This is the last version of the submitted manuscript, which is published in the international

academic journal ‘Public Management Review’:

Tummers, L.G., Vermeeren, B., Steijn, A.J. & Bekkers, V.J.J.M. (2012). Public

professionals and policy implementation: conceptualizing and measuring three types of

role conflicts. Public Management Review, 14, 2, 1041-1059.

Erasmus University Rotterdam

Dept. of Public Administration

Corresponding author: Lars Tummers

Tummers@fsw.eur.nl

0
Abstract (max. 100 words, now 100 words)

Nowadays, public policies often focus on economic values, such as efficiency and

financial transparency. Public professionals often resist implementing such policies. We

analyse this using the concept of ‘role conflicts’. We use a novel approach by

conceptualizing and measuring role conflicts on the policy level, thereby linking policy

implementation and social psychology research. We construct and test scales for policy-

client, policy-professional and organizational-professional role conflicts. Using survey

data, we show that policy-professional and policy-client role conflicts negatively influence

the willingness of public professionals to implement policies. Concluding, we

conceptualized and measured three role conflicts that can occur during policy

implementation.

Key words

 Role conflicts

 Policy implementation

 Public professionals

 Resistance to change

 Scale development

1
1 Introduction

Within the public administration literature, there is an intense debate concerning the

pressures that public professionals face when implementing public policies (Ackroyd et al.,

2007; De Ruyter et al., 2008; Freidson, 2001). It seems that many public professionals are

unwilling to implement public policies laid down by the government (Duyvendak et al.,

2006; Hebson et al., 2003). For instance, in the Netherlands, many insurance doctors

encountered substantial professional and moral concerns when asked to implement a new

policy focused on re-examining welfare clients. In fact, about 240 doctors urged a strike

against this new policy, and some decided to simply quit their job (Tummers et al., 2009).

Other examples from Canada show that public professionals often do not accept new

policies, and sometimes leave and start their own organizations (White, 1996).

When public professionals are unwilling to implement public policies, serious

consequences can result. First, it can significantly decrease the effectiveness of policy

implementation (Ewalt & Jennings, 2004; May & Winter, 2009). Second, the quality of

interactions between professionals and citizens may be affected, possibly influencing the

output legitimacy of government (Bekkers et al., 2007).

One important factor influencing the willingness to implement public policies seems

to be the conflicts that professionals experience during policy implementation. Many

contemporary policies focus strongly on economic values, such as efficiency and financial

transparency. This can be seen as an outcome of the influence of New Public

Management (NPM) (Hood, 1991). Public professionals may have difficulty in accepting

the changing trade-offs in values – due to the introduction of NPM reforms – which

become manifest when implementing a policy programme (Duyvendak et al., 2006;

Freidson, 2001).

These difficulties that professionals experience during the implementation of NPM

policies can be understood using the concept of role conflicts, as developed within the

2
social psychology literature (Kahn et al., 1964; Rizzo et al., 1970). When implementing a

policy, professionals face different demands from a range of role providers. Role conflicts

arise when professionals perceive these demands to be incompatible. Professionals

working at the front-line experience a number of role conflict types (Lipsky, 1980:46). For

example, a policy-client role conflict occurs when professionals perceive that the behavior

demanded by the policy they have to implement (such as following strict policy rules) is

incompatible with the behavior demanded by their clients (who want their situation to be

taken into account). It seems that the introduction of NPM policies has increased the

number of role conflicts as the values behind these policies (such as efficiency) can run

counter to professional values (such as equity) (Duyvendak et al., 2006; Freidson, 2001).

In this study, we aim to examine the influence of role conflicts on the willingness to

implement public policies. Our main research question is therefore:

What is the influence of the role conflicts encountered by public professionals during

policy implementation on their willingness to implement public policy?

To be able to answer this research question, we firstly conceptualize and measure the

role conflicts that occur during policy implementation, thereby combining insights from

both the policy implementation and the social psychology literatures. Numerous authors

have stressed the perverse effects of such conflicts (for example Honig, 2006;

Noordegraaf & Steijn, forthcoming 2011; Schneider, 1982; Tummers et al., 2009) using

qualitative case studies as the basis for their conclusions. In this study, a novel approach

is used by quantitatively studying role conflicts during policy implementation. Here, we use

a novel approach by conceptualizing and measuring role conflicts on the policy level,

thereby linking policy implementation and social psychology research. Scale development

techniques are used to conceptualize and measure these conflicts. This is valuable since,

3
 as Harris (1991: 125) notes, ‘further theoretical explication and scale construction is

necessary for research to proceed in this area. Researchers are encouraged to develop

and use more specific subscales to measure role ambiguity and role conflict in future

studies.’

After conceptualizing and measuring role conflicts on the policy level, we will

examine their effects on the willingness of public professionals to implement public policy.

In previous research, role conflicts have been related to job-level indicators such as

stress, burnout, poor life satisfaction, difficulty in decision-making and poor job

performance (Jackson & Schuler, 1985; Netemeyer et al., 1990; Tubre & Collins, 2000).

However, in this research, we focus on the policy-level effects of role conflicts by looking

at professionals’ willingness to implement a policy.

This brings us to the outline of this article. In Section 2, we consider the theoretical

framework by relating literature on role conflicts and on policy implementation to examine

the role conflicts faced by public professionals during policy implementation. In Section 3,

our method for measuring role conflicts and testing the hypotheses are outlined. The

results are presented in Section 4. We conclude the article by discussing the contribution

this research makes to the policy implementation literature and the debate on

professionals in public service delivery.

2 Theoretical framework

2.1 Introducing role conflicts

Organizational roles and role conflicts have been studied for at least sixty years (Tubre &

Collins, 2000), starting with the work of key scholars such as Merton (1949) and Parsons

(1951). Further, social psychologists such as Kahn et al. (1964) and Rizzo et al. (1970)

have extensively studied role conflicts.

4
 Examining the role conflict concept in work environments, Tubre & Collins (2000:

156) note that, ‘conceptually, a role is a pattern of behaviours perceived by an employee

as behaviours that are expected’. That is, employees often base their perceptions of the

duties and expectations associated with their chosen profession on the definition of jobs

and roles that others communicate to them (Kahn et al., 1964). A lack of compatibility

between multiple expectations can create conflict and tension. When people are

confronted with contradictory and competing role expectations, a situation described in

occupational stress research as a role conflict arises. More specifically, Katz and Kahn

(1978: 204) define a role conflict as ‘the simultaneous occurrence of two or more role

expectations such that compliance with one would make compliance with the other more

difficult.’

An extensive body of knowledge concerning role conflicts has been built up.

Research on role conflicts has been reviewed in three meta-analyses (Fisher & Gitelson,

1983; Jackson & Schuler, 1985; Tubre & Collins, 2000). In general, role conflicts have

been linked to a range of negative job attitudes and behaviours (Jackson & Schuler, 1985;

Netemeyer et al., 1990; Schaubroeck et al., 1998; Tubre & Collins, 2000). Nevertheless,

some more positive effects have been recorded. For instance, Lowenthal et al. (1975:110)

note that when individuals experience conflicting demands, this can increase opportunities

to develop a more distinct personality at later points in life.

2.2 Applying role-conflict ideas to policy-implementing public professionals

The degree and type of role conflicts that arise during policy implementation can depend

on the type of policy and the policy process. Wilson (2005) makes distinctions between

redistributive policy, protective and competitive policy, morality policy and distributive

policy. Within an NPM context, the government tries to enhance public performance by

introducing more businesslike practices within the public sector. In this respect, NPM

5
policies focus strongly on economic values, such as efficiency and client choice (Hood,

1991). This type of policy is often top-down (Hill & Hupe, 2009) in nature and often fails to

deliver tangible benefits to the regulated (Wilson, 2005). As noted in the introduction, such

policies can generate role conflicts for the implementing public professionals (see also

Duyvendak et al., 2006; Farrell & Morris, 2003; Lipsky, 1980; Smullen, forthcoming 2011;

Tummers et al., 2009).

Based on literature from the sociology of professions (especially Freidson, 2001)

and policy implementation streams (Duyvendak et al., 2006; Lipsky, 1980; Tummers et al.,

2009), we have identified three role conflicts which are considered especially important

when implementing public policies in such a situation: a policy-professional role conflict, a

policy-client role conflict and an organizational-professional role conflict.

Firstly, we would argue that a policy-professional conflict can occur. Policy

requirements are reflected in the policy contents, which are often laid down in formal rules

and regulations, such as the policy goals to be achieved. The role behaviour demanded

by these policy requirements can conflict with the professional values - the set of rules

one would follow if allowed to act professionally as a member of a professional

community. A policy-professional role conflict occurs when professionals tasked with

implementing a policy perceive the role requirements demanded by the policy contents to

be incongruent their professional attitudes, values and behaviour. This conflict can be

particularly pronounced if the policy is implemented in a top-down way, without consulting

the professionals (Hill & Hupe, 2009).

The second type considered relevant is the policy-client conflict. This type of

conflict occurs when professionals tasked with implementing a policy perceive the role

behaviour demanded by their clients to be incongruent with the role behaviour demanded

by the policy content. For a somewhat extreme example, consider police officers who

have to implement stricter law enforcement, such as zero-tolerance, policies. As Lipsky

6
(1980: 47) notes, these police officers ‘must enforce laws they did not make in

communities where demands for law enforcement vary with the laws and the various

strata of the population’. As a result, enforcing zero-tolerance, in line with the policy

content, can conflict strongly with the role behaviour demanded by clients, who want

police officers to take account of their specific circumstances.

Thirdly, we distinguish the organizational-professional conflict. An organizational-

professional role conflict occurs when a professional tasked with implementing a policy

perceives the role behaviour demanded by the organization regarding policy

implementation to be incongruent with his or her professional attitudes, values and

behaviour. Although this is related to the policy-professional role conflict, it is logically

independent as it looks at the way the organization implements the policy, not at the policy

content itself. In public organizations, managers are important actors in the

implementation of policies. There may be conflicts between the role behaviour demanded

by these managers, which for instance may stress efficiency and focusing on quantifiable

targets during policy implementation, and professional values and attitudes. A good

example is of an insurance physician implementing a new policy on re-examining welfare

clients. He stated that ‘there is clearly a culture of repression. Management does not

understand that doctors need time. Tensions arise when doctors want to work accurately

and managers tell them that they have to do fifteen re-examinations a week’ (cited in

Tummers et al., 2009: 701). However, the overall picture may well be more mixed - that

conflicts do occur, but only in limited ways. Many managers may well have values that are

not that dissimilar from the professionals working under them, for example because they

are, or used to be, frontline professionals themselves (Thomas & Davies, 2005).

7
2.3 Role conflicts and willingness to implement policy

Having conceptualized three types of role conflicts, we can now examine the possible

consequences of these role conflicts on professionals’ willingness to implement public

policy. Many scholars see the commitment of public professionals as a prerequisite for the

effective implementation of a public policy (Ewalt & Jennings, 2004; May & Winter, 2009;

Van Meter & Van Horn, 1975). Further, it has been fairly consistently claimed within the

field of change management that a crucial condition for success is that employees are

willing to implement an intended change (Judson, 1991; Lewin, 1951). According to

planned change theories, an absence of this willingness will result in a situation where top

management's intentions to instil a change will not be transformed into real change efforts

by lower echelons (Judson, 1991). According to the ‘emergent change’ school of thought,

unwillingness will impede the process of small, bottom-up modifications, such that these

will no longer accumulate and amplify (Weick, 2000). Alongside this, the notions of

working, shirking or sabotage, as discussed by Brehm and Gates (1997), are relevant in

this context. When public professionals are unwilling to implement the policy, ‘shirking’ or

‘sabotage’ are more likely to occur than ‘working’, and this is undesirable in terms of policy

performance.

Although some prominent policy implementation scholars have emphasized the

crucial role of front-line professionals being willing to implement the policy (Ewalt &

Jennings, 2004; May & Winter, 2009), a validated scale for measuring this has not been

developed. Therefore, we will draw on the change management literature, which has a

long history of examining willingness/resistance to changes (Judson, 1991; Lewin, 1951),

and use the concept of change willingness that has been validated by Metselaar (1997).

Change willingness is defined as 'a positive behavioural intention towards the

implementation of modifications in an organization's structure, or work and administrative

processes, resulting in efforts from the organization member's side to support or enhance

8
the change process' (Metselaar, 1997: 42). In this article, the change refers to the policy

that the professionals are required to implement.

We can now relate the willingness to implement a public policy with the three types

of role conflict discerned above. Kahn et al. (1964) argue that role conflicts result in stress

and anxiety on the part of organizational members. Hamner and Tosi (1974:479) note that

‘it appears that role ambiguity and role conflict result in undesirable consequences for

organization members.’ (see also Tubre & Collins, 2000:156). We therefore expect

professionals to prefer policies with less conflict to policies with greater role conflict, and

that this will make them reluctant to implement policies of the latter variety. For instance,

we expect professionals who experience a strong policy-client role conflict during policy

implementation to be reluctant to implement the policy (Lipsky, 1980). We therefore

hypothesize that:

H1: The more professionals experience a policy-professional conflict, the more unwilling

they will be to implement a policy.

H2: The more professionals experience a policy-client conflict, the more unwilling they will

be to implement a policy.

H3: The more professionals experience a organizational-professional conflict, the more

unwilling they will be to implement a policy.

Although we expect intense role conflicts to negatively influence willingness to implement

a policy, we do not expect all role conflicts to have the same magnitude of impact on the

willingness to implement. First, a role conflict involving the policy (policy-professional or

policy-client) could be more important than an organizational-professional conflict. This

idea follows from one of the main conclusions of the work by Brehm and Gates, who note

that policy preferences can be more important in influencing the attitudes and behaviour of

9
 street-level bureaucrats than the coercive capacities of managers (1997:199; see also

Golden, 2000).

Further, Dias and Maynard-Moody (2007) note that a ‘social work narrative’ is

experienced by some frontline workers who focus on helping clients to achieve long-term

success. Such public professionals want to enhance their clients’ lives when implementing

a policy (Maynard-Moody & Musheno, 2003). If one accepts this social work narrative,

then it seems particularly important to consider the policy-client conflict above the

organizational-professional conflict. This also corresponds with the findings of Riccucci

(2005: 102) who concludes that ‘state-level as well as agency officials may have very little

influence over worker discretion. Rather, the reference point for workers’ discretionary

judgment may be the client.’

Summarizing, it seems that role conflicts can have different impacts on the

willingness to implement a policy. In the empirical analyses, we will therefore analyse and

try to explain the different impacts found in the case studied.

3 Method

3.1 The introduction of a new policy

To test the proposed relationships between role conflicts and the willingness to implement

public policy, we surveyed Dutch mental healthcare professionals responsible for

implementing a new reimbursement policy known as Diagnosis Related Groups (DRGs: in

Dutch Diagnose Behandeling Combinaties, DBC’s). These DRGs were introduced in the

Netherlands as part of the Health Insurance Law in 2008. This new Health Insurance Law,

and the associated DRGs, can be seen as an example of introducing regulated

competition into Dutch healthcare, a move in line with New Public Management ideas

(Hood, 1991).

10
 The three interrelated policy goals of the DRG policy are a) to increase

transparency in the costs and quality of healthcare, b) to increase efficiency and c) to

increase patient choice. Policymakers expected that, initially, when healthcare providers

start to use these DRGs (‘healthcare products’), the activities of healthcare providers

would become more transparent and hence comparable. After a sufficient amount of

transparency was achieved, health insurers and healthcare providers would then have to

negotiate the volume and price of DRGs. This should increase competition and thus

efficiency (which Hood (1991:5) also sees as a doctrinal component of NPM).

Furthermore, patients should be able to choose from among healthcare providers, thus

increasing patient choice and efficiency.

We used the DRG policy for three reasons. Firstly, public professionals - here

psychotherapists, psychologists and psychiatrists – have to implement this policy, which

makes it relevant to the debate on the pressures public professionals face in service

delivery. Secondly, the DRG policy focuses strongly on economic goals, such as

efficiency and financial transparency. It is recognized that policies that pursue these kinds

of goals are likely to arouse conflicts in professionals. Thirdly, in numerous countries,

there have been moves towards similar healthcare payment systems, including the USA,

Australia, Germany, England, Japan, Sweden and Belgium (Kimberly et al., 2009). The

widespread use of such policies increases the possibility of generalizing our eventual

conclusions.

3.2 Sampling and response rate

Our base sample consisted of 1800 mental healthcare professionals randomly selected

from the databases of two nationwide mental healthcare associations. We received 478

returns of our questionnaire, an effective response rate of 26%. We carried out a non-

response check, but this did not indicate that there was a bias in the respondents. The

11
 most common reason for non-response was that the targeted professional did not work

with DRGs (the reason given by 157 non-responding professionals), sometimes because

they had yet to be implemented in their organisations. The next most frequent reason

given was that the respondents had retired or changed occupation (17).

We chose to include only respondents who had answered all the questions,

resulting in an effective sample of 411 respondents. Of the valid respondents, 121 (29%)

were men and 290 (71%) women. This ratio is consistent with Dutch mental healthcare

professionals as a whole, with Palm (2008) noting that females made up 69% of this

workforce. The respondents’ average age was 48, which is slightly higher than the Dutch

national average for mental healthcare professionals (M = 44). Given the large number of

respondents, the similarity of the respondents with the wider population in terms of

demographic variables plus the results of the non-response check, we are confident that

our respondents are representative of the population.

3.3 Measuring role conflicts: item generation and expert review

In this study, we have used a novel approach by studying the role-conflict concept within

the domain of a specific public policy. As such, we had to develop a valid and reliable

measure for this concept. In so doing, we followed the recommendations for scale

development by DeVellis (2003).

Firstly, for each role conflict, ten items were generated, formatted as five-point Likert

scales. We further used templates in constructing these items since these allow the

researcher to use specific phrases that fit the context of the research (DeVellis, 2003: 62).

For example, instead of stating ‘the policy’ and ‘professionals’, the researcher can

rephrase these items using the specific policy and group of professionals that are being

examined, here ‘the DRG policy’ and ‘mental healthcare professionals’. This makes it

easier for respondents to understand the items, as they are tailored to the context, which

12
increases reliability and content validity (DeVellis, 2003). Further, by explicitly developing

templates, the developed items can be used more easily in other contexts (with other

policies and professionals). As an example, one of the general items for the policy-client

role conflict was:

I feel that I sometimes have to choose between the wishes of my clients and the rules

and regulations of the policy.

When it comes to mental healthcare professionals implementing the DRG policy, the item

becomes:

I feel that I sometimes have to choose between the wishes of my patients and the rules

and regulations of the DRG policy.

Secondly, to further increase content validity, 21 expert reviewers examined the initial pool

of items. These experts were selected for their various areas of expertise (DeVellis, 2003:

75) and included one specialist in electronic surveys, three experts on role conflicts, four

quantitative methodologists, five mental healthcare specialists and eight public

administration scholars.

After each expert interview, we considered adding or discarding certain items,

based on the comments received. Eventually, we chose the six items which were

considered best (on average by the respondents) for each role conflict to construct a pool

of items to be administered in the wider questionnaire. By selecting six items for each role

conflict type, we retained the possibility of deleting further items in later stages of the scale

development process (DeVellis, 2003: 57). We confirmed the validity of the final pool of

items by discussing it with three public administration experts, two quantitative

methodologists and one specialist in mental healthcare drawn from our original advisors.

13
3.4 Measuring role conflicts

We conducted an exploratory factor analysis in order to examine whether it was possible

to actually discern three different role conflicts based on the items used to measure them.

At this early stage in developing role-conflict types, exploratory factor analysis was

favoured over methods which test hypothesized groups (i.e. the three different role

conflicts), such as confirmatory factor analysis. As criteria for item deletion, we adopted

common statistical warning signs, such as items having correlations less than .40 or more

than .90 with other items in the dimension, items loading more than .30 on to two factors

(or with low overall communalities) and items having a negative contribution to Cronbach’s

alphas (Field, 2005; Hinkin, 1998).

Prior to conducting a factor analysis for the role conflicts, the inter-item correlation

matrix was examined. If our items truly measure the same underlying dimension (i.e., role

conflicts), then we would expect them to be related to each other. On this basis, we

examined the items and then deleted one that had initially been designed to measure the

policy-professional role conflict. This item’s correlation was less than .40 with the other

items and, on closer consideration, seemed more closely related to a factor linked to

influencing a role conflict than being an item measuring a role conflict.

In our final exploratory factor analysis, three factors could be clearly identified,

based on the obtained scree plot, the Kaiser’s criterion and the theoretical meaningfulness

of the factors (DeVellis, 2003). The initial factor solution contained additional factors but

three of the items that were designed to measure the policy-client conflict had low

communalities. After deleting these ‘low-communality’ items, a clear three-factor solution

was produced.

Having identified the factor structure, we proceeded to determine the Cronbach’s

alphas for the three scales, a measure based on the correlations between the items, to

check for sufficient internal consistency. The alphas for the role-conflict scales were all

14
 above the minimum acceptable level of .70 (.81, .78 and .86 respectively). The results of

these analyses are presented in Table 1.

Table 1 Factor loadings for the final pool of items (pattern matrix)

Item Factor (F) loadings

F1 F2 F3
Policy-professional role conflict – eigenvalue 1.4, 10.0% variance explained
Looking from my professional values and norms, I embrace the policy. (R) .51 (.15) (.03)
The policy negatively affects my professional autonomy. .77 (.-14) (.09)
I have the feeling that I sometimes have to choose between my professional values and the .72 (.-12) (.21)
rules of the policy.
In working with the policy, I violate my professional ethics. .60 (.31) (.-06)
Working with the policy conflicts with my values and norms as a professional. .83 (.15) (.-06)

Policy-client role conflict – eigenvalue 1.0, 7.3% variance explained

Many of my clients complain to me about the policy. (.09) .84 (.09)
Working with the policy clashes with the wishes of many clients. (.17) .56 (.14)
My clients experience the policy as a breach of their privacy. (.14) .83 (-.02)

Organizational-professional role conflict – eigenvalue 6.2, 44.2% variance explained

Looking from my professional values and norms, I embrace the way my organization (.11) (-.05) .71
implemented the policy. (R)
The way my organization works with the policy conflicts with my professional autonomy. (.10) (.02) .79
I have the feeling that I sometimes have to choose between the way my organization (.21) (.02) .69
implements the policy and my professional values.
Exactly following my organization’s rules regarding the policy is incompatible with my (.20) (.09) .62
professional values.
I have professional concerns about the software systems my organization uses for the policy. (.06) (.25) .54
The way my organization handles the policy clashes with my norms and values as a (.09) (-.03) .87
professional.

Template words are indicated using underline type. These are policy (for the research ‘DRG policy’ was used), clients
(‘patients’ was used) and professional (‘healthcare professional’ was used). We discussed this with the five mental healthcare
specialists, and they were satisfied with using these template words.

3.5 Measuring willingness and control variables

We measured willingness to implement the policy using the validated five-item scale of

Metselaar (1997). This scale uses templates to specify the change. Sample items are: ‘I

am willing to contribute to the introduction of DRGs’ and ‘I am willing to free up time to

15
 implement the DRG policy’. The answers were given on a five-point Likert scale. The

scale’s Cronbach’s alpha in this study was .85.

Finally, we included some control variables. Age (open question) was controlled for

on the basis of earlier findings that older employees tend to be less positive about change

and new policies (Wanberg & Banas, 2000). We also controlled for whether the

respondent was a manager of other professionals (0 = no, 1 = yes) given that managers

may have greater opportunities for participation, making them more willing to implement

the policy. Gender was also tested for its potential relevance. Further, we examined the

context in which the professional worked: as a freelance or in an institution. These types

of work are clearly different and could affect a professional’s experiences. Finally, the

occupation of the professional was examined, since the policy could have different

consequences for the diverse professions1.

4 Results

4.1 Descriptive statistics

Descriptive statistics and intercorrelations of the variables are presented in Table 2:

1
A professional can have multiple professions, such as being both a psychologist and a psychotherapist. In

Table 2, a professional’s responses are included under all the professions they claimed to belong to. No

reference category was constructed given this complexity.

16
Table 2 Descriptive statistics and correlations for the variables in the study

Variable Mean SD 1 2 3 4 5 6 7 8 9 10 11 12

1. Sex 0.71 NA

2. Age 47.97 10.89 -.24**

3. Occupation researcher (among 0.04 NA -.04 -.24**

else)

4. Occupation psychologist 0.74 NA .19** -.07 -.31**

(among else)

5. Occupation psychotherapist 0.45 NA -.05 .43** -.18** .16**

(among else)

6. Occupation psychiatrist (among 0.17 NA -.22** .12* -.06 -.74** -.28**

else)

7. Professional works only as a 0.24 NA .00 .37** -.11* .01 .40** -.09

freelance

8. Professional works in an 0.18 NA -.09 .16** -.09 -.03 .06 .10* -.26**

organization and as a freelance

9. Managing position 0.27 NA -.17** .09 -.07 -.19** -.10* .29** -.16** .12*

10.Policy-professional Role 3.72 0.77 -.07 .20** -.03 -.02 .22** .06 .10* -.02 -.05

Conflict

11. Policy- client RC 2.95 0.78 -.04 .22** -.05 -.04 .15** .07 .18** -.07 .00 .57**

12.Organization-Professional RC 3.25 0.73 -.09 .18** .06 -.12* .07 .10* -.01 .02 -.01 .64** .50**

13.Willingness to implement the 2.53 0.81 .13* -.18* -.04 .08 -.09 -.14* -.08 .01 .08 -.53** -.41** -.37**

policy

Note. *p < .05, **p < .01

17
 As can be seen from Table 2, all the bivariate hypothesized correlations of the variables

were statistically significant and in the anticipated direction.

The use of self-reported data from a single questionnaire can create distortions in

the data, in particular as a result of common method bias (Podsakoff & Organ, 1986). Of

concern here is that respondents were asked to rate both their degree of role conflicts as

well as their willingness to implement a policy. Any correlation between the policy-

professional role conflict, for example, and their willingness to implement the policy may

be attributable to the fact that employees were asked to rate both aspects within a single

questionnaire. To determine whether this was a valid concern, we conducted a Harman

one-factor test. A factor analysis was conducted on all the items. The factors together

accounted for 64% of the total variance (using the eigenvalue greater than unity criterion).

Further, the first (i.e. largest) factor did not account for a majority of the variance (only

27%). Given that no single factor emerged and that the first factor did not account for a

majority of the variance, common method variance does not seem to be a significant

concern with the results obtained.

4.2 Regression results

Hierarchical multiple regression analyses were conducted to examine the extent to which

the three role conflict types predicted willingness to implement the policy. Firstly, we

regressed willingness to implement the policy on the control variables. Next, we added the

policy-client role conflict (model 2), the policy-professional role conflict (model 3), and the

organizational-professional role conflict (model 4). In each step, the change in explained

variance (R2) was calculated, and we determined whether each change was significantly

different from zero.

In the first model, with control variables in the equation, the R2 was .08 (F=4.07,

p<.01). Inclusion of the policy-professional role-conflict scores in the second model

18
increased the R2 to .32. After adding the policy-client role-conflict scores, the R2 increased

to .33. Finally, with the insertion of the organizational-professional role conflict, the R2

remained at .33. Thus, the combination of the three role conflicts contributed considerably

to the implementation (un)willingness experienced by public professionals. We can now

look at the specific hypotheses.

Hypothesis 1 predicts that the degree of policy-professional role conflict

experienced by public professionals will be negatively related to their willingness to

implement DRGs. As Table 3 shows, when we look at the final model, the policy-

professional role conflict was indeed significantly related to willingness to implement the

policy (β=-.42, p<.01). That is, when public professionals feel that the policy is in conflict

with their professional values, they are less inclined to make an effort to implement the

policy. With respect to Hypothesis 2, the results also show a significant negative

relationship between policy-client conflict and implementation willingness (β=-.15, p<.01).

So, when professionals experience a stronger conflict between policy demands and the

demands and wishes of their clients, they become less willing to implement the policy.

Hypothesis 3 predicts that a more intense organizational-professional role conflict reduces

willingness to implement a policy. However, this relationship appeared not to be significant

once all the other factors were controlled for. Therefore, this hypothesis is rejected.2

These results support our expectation that role conflicts do not all have the same

magnitude of impact on willingness to implement a policy. However, we had not expected

the organizational-professional role conflict to have no effect at all on the willingness to

implement a policy. We were able to further examine this insignificant influence of the

2
Although the correlation of this role conflict with willingness to implement a policy was statistically

significant, in the regression analysis this influence became insignificant once other variables were

taken into consideration. This is not due to multicollinearity, which is sometimes the cause, since

this was not an issue in our regression, as shown by the low VIF values.

19
organizational-professional role conflict. Based on expert interviews and numerous open

answers in our survey, we concluded that one major reason for this finding is that the

organizations had very little influence on the way the DRG policy was implemented as the

rules were rather stringent. Professionals stated that ‘My organization had little choice

other than to go along with the national DRG policy’ and ‘My impression is that my

organization had few options in the execution of the DRG policy’.3 Further, some

managers shared the values of professionals, often not being that fond of the DRG policy

themselves. As one professional noted: ‘I think that the DRG policy was for organizations -

including my own - a necessary evil. They had to make the best of it. Nobody is happy

about this.’ One of the few things the organization did have an influence on was the ICT

system used for the DRG policy, but this was not significant compared to the changes

brought about by the DRG policy itself.

3
Quotations are drawn from open answers recorded in the survey described in this article and

translated from the Dutch originals.

20
Table 3 Hierarchical regression analyses for variables predicting willingness to implement the policy

Model 1 – including Model 2 – including Model 3 – including Model 4 – including

control variables policy-professional policy-client organizational-
role conflict role conflict professional
role conflict

Woman .07 .06 .06 .06

Man Ref.cat. Ref.cat. Ref.cat. Ref.cat.

Age -.15** -.09 -.08 -.08

Occupation: researcher -.15** -.11** -.11* -.11*

Occupation: psychologist -.18* -.11 -.11 -.11

Occupation:
psychotherapist -.09 -.02 .02 .02

Occupation: psychiatrist -.30** -.18** -.18* -.18*

Professional works only as

.01 .01 .00 .00
freelance.
Professional works as
freelance and in an .04 .01 .00 .00
organization.
Professional works only in Ref.cat. Ref.cat. Ref.cat.
an organization. Ref.cat.

Managing position .13** .09** .09* .09*

Non-managing position Ref.cat. Ref.cat. Ref.cat. Ref.cat.

Policy-professional role
-.50** -.42* -.42*
conflict

Policy-client role conflict -.15** -.15**

Organizational- .00
professional role conflict

ΔR2 .23** .02** 00

Overall R2 .08** .32** .33** .33**

Note: Standardized coefficients are presented. * p < .05 ** p < .01.

The following criteria are met (based on Field, 2005):

Criterion of independent residuals (Durbin-Watson 1.8, 1<criterion<3). Criterion for no multicollinearity (all VIF values below
10 and average close to 1). No exclusion of influential outlying cases was required (using casewise diagnostics, 4.6% above
standardized residual >|2|), Cook’s distance max. 0.09 (criterion < 1). Criteria for homoscedasticity and normality met.

21
5 Discussion and conclusions

Public professionals are often unwilling to implement public policies, especially when

these policies focus on business values such as efficiency (Duyvendak et al., 2006;

Freidson, 2001; Noordegraaf & Steijn, forthcoming 2011). This can be an undesirable

situation given that policy implementation scholars have shown that implementers’

willingness to implement public policies is crucial for policy performance (Ewalt &

Jennings, 2004; May, 2003). In this study, we have aimed to quantitatively examine the

impact of role conflicts on this (un)willingness to implement public policies. Based on our

analyses, we can draw three conclusions that should be of interest to policy

implementation scholars and practitioners.

Firstly, we have added to the policy implementation literature by conceptualizing

and measuring three role conflicts that could occur during policy implementation.

Following the recommendations for scale development by DeVellis (2003), we developed

corresponding role conflict scales. Other researchers can use these scales to examine

role conflicts during policy implementation. Policymakers and managers could use these

scales to discover which conflicts are occurring during the implementation of a particular

policy. Based on the results, they could try to alleviate the most intense or influential role

conflicts.

Secondly, using a survey of 411 mental healthcare professionals, we showed that

when professionals experience role conflicts, they are indeed less willing to implement

new policy programmes. The three role conflict types, together with conventional control

variables, explain over 30% of the variance in willingness to implement the policy. Looking

at the relatively large impact of role conflicts on willingness to implement policy, we

therefore recommend scholars and practitioners to include the role conflict concept when

studying policy implementation on the street-level.

22
 A related future research suggestion is to quantitatively examine the relationships

among role conflicts, policy performance and discretion – something we did not explore

here. On the one hand, discretion could enable professionals to cope more effectively with

role conflicts, and this should therefore increase their willingness to implement a policy

(and therefore boost policy performance) (see also Maynard-Moody & Musheno, 2000).

On the other hand, more discretion could exaggerate the implementation gap. Exploring

these relationships could provide new insights into the experiences and behaviour of

frontline professionals during policy implementation.

Our third conclusion relates to the individual role conflict types. The policy-client

and the policy-professional conflicts proved to be influential in explaining willingness to

implement policies. In the case studied, the policy-professional role conflict proved the

most important: implementers who could not align their professional values with the rules

of the policy, were far less willing to implement it. This highlights the importance of

professional values during policy implementation, and their potential conflicts with a new

policy (see also Anderson, 2010). This is an important conclusion for policymakers and

managers. They need to be aware of this conflict when a policy is to be implemented by

professionals such as teachers, physicians or psychologists. Policy implementation

scholars could further examine this role conflict, and they could use insights from the

sociology of professions literature which has a long history of examining professionalism

and how it conflicts with other ‘value systems’ (Eraut, 1994; Freidson, 2001).

Unlike the policy-professional and the policy-client conflicts, the organizational-

professional role conflict did not appear to be a significant determinant of the willingness

to implement a governmental policy in the case studied. One major reason for this is, with

respect to the DRG policy, that the policy rules were rather strict, making it difficult for

organizations to adapt the policy to their situation. This conclusion nuances the often

stated manager-professional clash (Davies & Harrison, 2003; Wilensky, 1964). It is more

23
in line with the view of those scholars who note that many managers may well have

values and goals that are not that dissimilar from those of professionals (Crilly & Le

Grand, 2004; Hewison, 2002) and that managers are increasingly subject to forces

beyond their own organization, such as new policies, politics and the media (Noordegraaf

& Steijn, forthcoming 2011). In line with this argument, we found that, in our case,

managers had to implement policies that they did not in principle agree with, and ‘sell’

these policies within their organizations. Policy implementation scholars could usefully

further research and highlight this intriguing managerial role during policy implementation.

As with all studies, this one had a number of limitations. Here, we discuss two

important limitations. Firstly, the results of this study, and the implications drawn, should

be interpreted in light of the study's limited context and sample. Although the study's

generalizability was increased by the large number of public professionals involved, and

that these were working in different occupations, positions and places, one should be

cautious in generalising this to other public-sector policies or domains. An area for further

research would be to test the proposed model on other types of policies in a range of

public domains.

A second limitation of this study is that it did not explicitly take the organizational

context into account. Future studies could rectify this. For instance, what are the

characteristics of the implementing organization and the implementing professionals? A

number of interviews with specialists drawn from the specific field could be helpful before

conducting a survey. Furthermore, following a survey, it would seem worthwhile to discuss

its results with such specialists in order to contextualize the results.

Concluding, this study has conceptualized and measured three role conflicts that

public professionals can experience during policy implementation. Researchers can use

these instruments to measure role conflicts during policy implementation. Further, this

study has shown how these role conflicts impact on the willingness (or resistance) of

24
public professionals who have to implement public policies. Future research into the role

conflicts experienced by professionals during policy implementation, including ways to

minimize these conflicts can be a timely and productive endeavour for both researchers

and practitioners alike.

References

Ackroyd, S., Kirkpatrick, I., & Walker, R. M. (2007). Public management reform in the UK

and its consequences for professional organization: A comparative analysis. Public

Administration, 85(1), 9-26.

Anderson, J. E. (2010). Public policymaking: An introduction. Boston: Wadsworth.

Bekkers, V. J. J. M., Edwards, A., Fenger, M., & Dijkstra, G. (2007). Governance and the

democratic deficit, assessing the legitimacy of governance practices. Aldershot:

Ashgate.

Brehm, J., & Gates, S. (1997). Working, shirking, and sabotage: Bureaucratic response to

a democratic public. University of Michigan press: Ann Arbor, MI.

Crilly, T., & Le Grand, J. (2004). The motivation and behaviour of hospital trusts. Social

Science & Medicine, 58(10), 1809-1823.

Davies, H. T. O., & Harrison, S. (2003). Trends in doctor-manager relationships. BMJ,

326, 646-649.

De Ruyter, A., Kirkpatrick, I., Hoque, K., Lonsdale, C., & Malan, J. (2008). Agency working

and the degradation of public service employment: The case of nurses and social

workers. The International Journal of Human Resource Management, 19(3), 432-445.

DeVellis, R. F. (2003). Scale development: Theory and applications. Thousand Oaks:

Sage.

25
Dias, J. J., & Maynard-Moody, S. (2007). For-profit welfare: Contracts, conflicts, and the

performance paradox. Journal of Public Administration Research and Theory, 17(2),

189.

Duyvendak, J. W., Knijn, T., & Kremer, M. (Eds.). (2006). Policy, people, and the new

professional. de-professionalisation and re-professionalisation in care and welfare.

Amsterdam: Amsterdam University Press.

Eraut, M. (1994). Developing professional knowledge and competence. London:

Routledge.

Ewalt, J. A. G., & Jennings, E. T. (2004). Administration, governance, and policy tools in

welfare policy implementation. Public Administration Review, 64(4), 449-462.

Farrell, C., & Morris, J. (2003). The neo-bureaucratic'state: Professionals, managers and

professional managers in schools, general practices and social work. Organization,

10(1), 129-156.

Field, A. (2005). Discovering statistics using SPSS (2nd ed.). London: Sage.

Fisher, C. D., & Gitelson, R. (1983). A meta-analysis of the correlates of role conflict and

ambiguity. Journal of Applied Psychology, 68(2), 320.

Freidson, E. (2001). Professionalism: The third logic. Cambridge: Cambridge University

Press.

Golden, M. M. (2000). What motivates bureaucrats? politics and administration during the

reagan years. New York: Columbia University Press.

Hamner, W. C., & Tosi, H. L. (1974). Relationship of role conflict and role ambiguity to job

involvement measures. Journal of Applied Psychology, 59(4), 497-499.

Harris, M. M. (1991). Role conflict and role ambiguity as substance versus artifact: A

confirmatory factor analysis of house, schuler, and levanoni's (1983) scales. Journal

of Applied Psychology, 76(1), 122-126.

26
Hebson, G., Grimshaw, D., & Marchington, M. (2003). PPPs and the changing public

sector ethos: Case-study evidence from the health and local authority sectors. Work

Employment and Society, 17(3), 481-501.

Hewison, A. (2002). Managerial values and rationality in the UK national health service.

Public Management Review, 4(4), 549-579.

Hill, M., & Hupe, P. (2009). Implementing public policy (2nd ed.). Thousand Oaks: Sage.

Hinkin, T. R. (1998). A brief tutorial on the development of measures for use in survey

questionnaires. Organizational Research Methods, 1, 104-121.

Honig, M. I. (2006). Street-level bureaucracy revisited: Frontline district central-office

administrators as boundary spanners in education policy implementation. Educational

Evaluation and Policy Analysis, 28(4), 357-383.

Hood, C. (1991). A public management for all seasons. Public Administration, 19(1), 3-19.

Jackson, S. E., & Schuler, R. S. (1985). A meta-analysis and conceptual critique of

research on role ambiguity and role conflict in work settings. Organizational Behavior

and Human Decision Processes, 36(1), 16-78.

Judson, A. S. (1991). Changing behavior in organization: Minimizing resistance to change.

Cambridge, MA: Basil Blackwell.

Kahn, R. L., Wolfe, D. M., Quinn, R. P., & Snoek, J. D. (1964). Organizational stress:

Studies in role conflict and ambiguity. New York: Wiley.

Katz, D., & Kahn, R. L. (1978). The social psychology of organizations. New York: Wiley.

Kimberly, J. R., de Pouvourville, G., & Thomas, A. D. A. (2009). The globalization of

managerial innovation in health care. Cambridge: Cambridge University Press.

Lewin, K. (1951). Field theory in social science New York: Harper.

Lipsky, M. (1980). Street-level bureaucracy. New York: Russell Sage Foundation.

Lowenthal, M. F., Thurnher, M., & Chiriboga, D. A. (1975). Four stages of life Jossey-Bass

Publishers.

27
May, P. J. (2003). Policy design and implementation. In B. Guy Peters, & J. Pierre (Eds.),

Handbook of public administration (pp. 223-233). London: Sage.

May, P. J., & Winter, S. C. (2009). Politicians, managers, and street-level bureaucrats:

Influences on policy implementation. Journal of Public Administration Research and

Theory, 19(3), 453.

Maynard-Moody, S., & Musheno, M. (2000). State agent or citizen agent: Two narratives

of discretion. Journal of Public Administration Research and Theory, 10(2), 329.

Maynard-Moody, S., & Musheno, M. C. (2003). Cops, teachers, counselors: Stories from

the front lines of public service. University of Michigan: University of Michigan Press.

Merton, R. K. (1949). Social theory and social structure. Glencoe: Free Press.

Metselaar, E. E. (1997). Assessing the willingness to change: Construction and validation

of the DINAMO. (Doctoral dissertation, Free University of Amsterdam).

Netemeyer, R. G., Johnston, M. W., & Burton, S. (1990). Analysis of role conflict and role

ambiguity in a structural equations framework. Journal of Applied Psychology, 75(2),

148-157.

Noordegraaf, M., & Steijn, A. J. (Eds.). (forthcoming 2011). Professionals under pressure:

Perspectives on professionals and professionalism. Amsterdam: Amsterdam

University Press.

Palm, I., Leffers, F., Emons, T., Van Egmond, V., & Zeegers, S. (2008). De GGz

ontwricht: Een praktijkonderzoek naar de gevolgen van het nieuwe zorgstelsel in de

geestelijke gezondheidszorg. Den Haag: SP.

Parsons, T. (1951). The social system. Glencoe: Free Press.

Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research:

Problems and prospects. Journal of Management, 12(4), 531-544.

Riccucci, N. M. (2005). Street-level bureaucrats and intrastate variation in the

implementation of temporary assistance for needy families policies. Journal of Public

Administration Research and Theory, 15(1), 89-111.

28
Rizzo, J., House, R., & Lirtzman, S. (1970). Role conflict and ambiguity in complex

organizations. Administrative Science Quarterly, 21, 598-610.

Schaubroeck, J., Judge, T. A., & Taylor, L. A. (1998). Influences of trait negative affect

and situational similarity on correlation and convergence of work attitudes and job

stress perceptions across two jobs. Journal of Management, 24(4), 553-576.

Schneider, A. L. (1982). Studying policy implementation. Evaluation Review, 6(6), 715-

730.

Smullen, A. (forthcoming 2011). Institutionalizing professional conflicts through financial

reforms: The case of DBC's in dutch mental healhcare. In M. Noordegraaf, & A. J.

Steijn (Eds.), Professionals under pressure: Perspectives on professionals and

professionalism (). Amsterdam: Amsterdam University Press.

Thomas, R., & Davies, A. (2005). Theorizing the micro-politics of resistance: New public

management and managerial identities in the UK public services. Organization

Studies, 26(5), 683-706.

Tubre, T. C., & Collins, J. M. (2000). Jackson and schuler (1985) revisited: A meta-

analysis of the relationships between role ambiguity, role conflict, and job

performance. Journal of Management, 26(1), 155.

Tummers, L. G., Bekkers, V. J. J. M., & Steijn, A. J. (2009). Policy alienation of public

professionals: Application in a new public management context. Public Management

Review, 11(5), 685-706.

Van Meter, D. S., & Van Horn, C. E. (1975). The policy implementation process: A

conceptual framework. Administration & Society, 6(4), 445-488.

Wanberg, C. R., & Banas, J. T. (2000). Predictors and outcomes of openness to changes

in a reorganizing workplace. Journal of Applied Psychology, 85(1), 132-142.

Weick, K. E. (2000). Emergent change as a universal in organizations. In M. Beer, & N.

Nohria (Eds.), Breaking the code of change (pp. 223-241). Boston: Harvard Business

School Press.

29
White, D. (1996). A balancing act: Mental health policy-making in quebec. International

Journal of Law and Psychiatry, 19(3-4), 289-307.

Wilensky, H. L. (1964). The professionalization of everyone? American Journal of

Sociology, , 137-158.

Wilson, C. A. (2005). Public policy: Continuity and change. Boston: McGraw-Hill.

30